July 01, 2012

The Bronze Age Indo-European invasion of Europe

Last summer, I was eagerly awaiting the publication of the genome of the Tyrolean Iceman. It is quite remarkable that only a year later, there is now autosomal DNA from half a dozen prehistoric Europeans. By comparing their DNA to that of modern populations, we are beginning to understand how the current mosaic of European peoples was formed.

The ancient samples vary greatly in the number of SNPs tested, and we cannot be sure how well they map to the restricted range of modern populations. Nonetheless, a crystal clear general pattern seems to emerge, at least in its broadest outlines (Sources: Oetzi the Iceman, Neolithic Swedes, Mesolithic Iberians):


We can plainly see that European hunter-gatherers best map to the modern Atlantic_Baltic population component. This is well represented in the remotest areas of Europe, the ones most distant from the Near Eastern womb of nations. It can be reasonably supposed that the modern Atlantic_Baltic component partially captures alleles present in the ancient European hunter-gatherers, the mtDNA haplogroup U population that seems to stretch from Iberia to Siberia. (However, note, that this does not mean that the Atlantic_Baltic component represents hunter-gatherer ancestry only.)

Conversely, ancient European farmers also possess a large chunk of the Southern component which is absent in the hunter-gatherers. This occurs at high frequencies today around the Mediterranean and reaches its maximum in the Near East. It is clear that there is direct evidence that farming came to Europe not as an idea, but as a people, just as archaeology and physical anthropology had always indicated -- until the rather modern distaste for migration set in.

But there is another component present in modern Europe, the West_Asian which is conspicuous in its absence in all the ancient samples so far. This component reaches its highest occurrence in the highlands of West Asia, from Anatolia and the Caucasus all the way to the Indian subcontinent. It is well represented in modern Europeans, reaching its minima in the Iberian peninsula, Sardinia, and Finland. A sampling of populations, including those closest geographically to the ancient samples:


There are several observations we can make:
  • The West_Asian component has a pan-European distribution: it must have been involved in a pan-European process rather than a more localized historical phenomenon.
  • Its absence from prehistoric individuals down to ~5ky ago suggests that it may have been added to the European population at a later date, although it may already have been present in currently unsampled areas (e.g., the Balkans) prior to 5kya.
  • It reaches its lowest occurrence in areas where non-Indo-European languages have been spoken (Basques and Iberia in general, Sardinia, and Finland)
The post-5kya timeframe is also conventionally accepted by linguists for either the dispersal of Indo-European languages, or at least of a significant subset thereof.

In that respect, it is important to note the correspondence between the West_Asian autosomal component and the k5 component of Metspalu et al. (2011): the latter is the major West Eurasian element in the Indian subcontinent. If a major episode of West Eurasian admixture took place in India 1,200-4,000 years ago, and keeping in mind uncertainties about dating, it may very well be that this corresponds, at least in part, to the eastern manifestation of the same phenomenon.

The Bronze Age is an important transitional phase in European archaeology: distinctive archaeological cultures with distinctive physical types make their appearance across the continent. There appears to be substantial innovation in metallurgy, weapons and transportation, increase in raiding, abandonment of settlements, and formation of broad-range confederacies with distinctive archaeological markers.

I propose that a quantum leap in social complexity occurred during this period. Remarkably, after ~4,000YBP, there are no longer farmers and hunter-gatherers as distinct cultures anywhere in Europe, and their mtDNA gene pools begin to expand in synch with each other. It may very well be that climatic upheavals framing this period may have triggered population movements, both Indo-European and Semitic.


Perhaps, through a combination of better technology and social organization, the Indo-European speaking nucleus, originally one among many linguistic groups of the prehistoric Near East were able to transmit their language, culture, and ideology to much larger populations, by alternatively subjugating or incorporating them. We can thus view the Indo-European bearers as "creative destructors", upsetting the balance of established societies and re-creating them in their own image.

Both the wide differences in genetic composition among present-day Indo-European speakers, and their early-attested physical contrasts testify to the fact that the original IE nucleus did not maintain itself apart --at least not for long!-- from the populations they encountered; in this they appear different from the earlier farmers who apparently kept their Mediterranean affinity even in the northernmost edge of their expansion, thousands of years after their entry into Europe.

Nonetheless, some of the legacy of the earliest Indo-European speakers does appear to persist down to the present day in the genomes of their linguistic descendants, and I predict that when we sample later (post 5-4kya) individuals we will finally find the West_Asian piece that is missing from the European puzzle.


* * *
ADDENDUM:


On the right are the K12b results for the ancient individuals, which further suggest that the Mesolithic Europeans possessed a Northern European affinity that is lacking in the farmer individuals. Conversely, farmer individuals have an excess of the Atlantic_Med component, and also a presence of the Southwest_Asian one.

These results are compatible with the above discussion, as well as the observation by Sánchez-Quinto et al. (2012) that:
the position of La Braña individuals in the 1000 Genomes Project data and the 1KGPomnichip PCAs suggests that the uniform Mesolithic substrate could be related to modern Northern European populations but may represent a gene pool that is no longer present in contemporary Southern European populations.
I wrote a small survey a couple years ago, in which I argued that:
I will simply say that the recent results of Bramanti et al. for a U-dominated older mtDNA stratum in Central/North-eastern Europe can be reasonably extended to cover both North-western Europe and northern Eurasia up to Lake Baikal, the prehistoric limit between Caucasoids and Mongoloids.
The extension of U dominance to Iberia, as well as the detection of the North European autosomal signal in both Iberia as well as, presumably, Siberia, seems to make the case for the existence of this postulated zone pretty clear.

One could very well call this boreal population Ancestral Northwest Eurasians (ANWEA) and parallelize it with the Ancestral South Indians (ASI). These Paleolithic substrata frame the Eurasian landmass on opposite latitudinal ends, and were the receptacles of the great chain of migrations which began in the Neolithic womb of nations. One of the final episodes of this process was the dispersal of the Indo-European languages during the Copper and Bronze Ages.

UPDATE (July 3): Indo-European genetic signatures in an Orcadian genome and a Lithuanian and Ancient European DNA using DIYHarappaWorld and MDLP5

35 comments:

  1. Can anyone recommend some reading material that might help me understand all of this? I have a background in science (physicist) but all this makes very little sense to me. I hate to admit it but i ve been subscribed to this blog for a long time now and i cant really read the articles !

    ReplyDelete
  2. So, what exactly pulls those Spanish HGs to far into Asia on a PCA?

    Its not showing in these admixture components.

    ReplyDelete
  3. Kostas:

    First, the Ks. Here we have several thousand datapoints in a very high-dimensional vector space. Some groups of points are closer to each other than others, thus you can form "clusters". Inside those clusters you can find sub-clusters and so on.
    Sometimes it is useful to analyse data with 3 clusters, but in other occasions you'll need 10 or 20. "k" is the total number of clusters. When Dienekes says "in k 7" he means that he's splitting his gene database in 7 clusters, and then he studies which parts of the Tyrolean Iceman's genome fit better into each cluster. As you can see in the results, 51.9% of Oetzi the Iceman's genome "fits better" among the cluster of people from Southern Europe.

    You can picture it like this: the names "East Asian, Atlantic Baltic, etc" are analogous to the bases that generate a vector field.

    But, what is the ideal number of clusters? In computer science, as a rule, you have to minimize a function F = G + H, where G is the sum of the squeare distances between each point and its cluster's centre, and H is a penalty that grows polynomially with the number of clusters. Thus you keep a balance between the number of clusters and the accuracy.

    The huge and terrible problem here is that genetics use a computer program called ADMIXTURE, whose source code is unavailable. So for me, the exact procedure is just magic.



    Other than that, you'll have to learn lots of history and stuff, like all the pre-historical cultures of Europe etc.

    ReplyDelete
  4. Thanks Baldric

    That clears up the math part quite a bit even though i don't understand the bases analogy. It sounds more like these components are subspaces rather than the base of the space.

    But still if i am to make any real world sense out of it i need something about the physical interpretation. What are the bases of that space? What does the measurement of an individual look like (a certain number of points?) If this procedure gives you a number of clusters what makes you think that they have some physical interpretation? I can think of a hundred questions.

    Isnt there something that explains what is going on here on a basic level? I dont even need a book, just some webpage or something short and basic.

    ReplyDelete
  5. Is interesting to see that if you mate Gok4 with the Swedish hunters, for example Ajv52 it gives similar results to La Braña individual, for the major components :
    North-Euro : 5.5 + 77.3 / 2 = 41.5
    Atlanto-Med : 81 + 13.3 / 2 = 47.15

    ReplyDelete
  6. Kostas:

    Ancestral component analysis fits as well as it can a linear decomposition of a matrix of genetic data (tens of thousands of SNPs each) from a certain number of individuals (up to a few thousand) into K eigenvectors (where the value of K is chosen prior to doing the analysis and is typically performed for everything from two up to a dozen or more eigenvectors) which are mixed in varying proportions to recreate a matrix in any given individual. The eigenvectors are assumed to represent distinct historic or prehistoric waves of mass population migration which after admixture produce a stable modern population genetic mix that has reached fixation locally.

    One compoent represents the first wave of modern humans who replaced Neanderthals in Europe starting from ca. 40 thousand years ago (kya) with minor Neanderthal admixture at some point and was genetically stable to within a few thousand years of the Neolithic; despite the depopulation of Europe during the Last Glacial Maximum ca 20kya and its repopulation from several Southern Europe refugia. Ancient mtDNA suggests that Cro-Magnon populations were racially unified and very different from farmer migrants. In the Neolithic itself (or the Epipaleolithic a few thousand years earlier), new peoples arrive on the scene and ancient DNA evidence suggests that there was major population genetic change. But, early Neolithic genetics also don't look like modern European population genetics, so there were presumably one or more subsequent mass population shifts. At least one of those mass population impacts is the admixture of Indo-European conquering ("superstrate") populations into earlier European farming and herding populations. There may have been additional layers of genetic influence.

    Major controversy exists regarding the details of how much admixture came with which waves at what times. The areas associated with Mesolithic layers are those that were the last refugia of hunting gathering populations in Europe as farming and herding took over first in prime land for that and eventually in more marginal land until it was virtually gone.

    The Neolithic starts in the Near East ca. 8,000 BCE, spreads to most of Europe from about 6,000 BCE to 3,500 BCE in fits and starts (with a very weak hold in fishing economies around the Baltic Sea and in the Arctic). Indo-European influence in most of Europe traces to the Bronze Age (ca. 2500 BCE to 1200 BCE), or perhaps a bit older - between Bronze Age and Neolithic is the "Copper Age" (aka The Chalcolithic aka Eneolithic ca. 5th to 3rd millenia BCE) (as Dienekes notes), or perhaps early Iron Age (ca. 1200 BCE to ca. 500 BCE) although his placement of the source of IE expansion as Near Eastern rather than in the Pontic-Capsian steppe is a minority view, and IE languages are probably much older in their core area than much of Europe. In India, the Indo-Aryan invasion that putatively brough Sanskrit derived languages and gave rise to the Hindu religion from the NW ca. 2000-1500 BCE. In the Aegean, the replacement of Minoan culture with Mycenean Greek marks the approximate non-Indo-European to Indo-European transition. The Urnfield cultures were probably IE; the copper age Bell Beaker culture's affiliation is a controversial. Wikipedia searches on key terms are good ways to get a sense of it.

    "Guns, Germs and Steel" and "First Farmers" are some of the better major book length treatments of the main ideas.

    ReplyDelete
  7. The only thing I can't wrap my head around is prevalence of R1B in the Basque along with the prevalence of R1B in rest of Western Europe.

    Since it is looking as if R1B arrived late, during the bronze age with the Bell-Beaker culture, R1B is the most likely candidate signal for the Indo-European invaders. However it doesn't jibe because the Basque show high levels of R1B yet lack the West Asian component present with their R1B dominant, Indo-European speaking neighbors (ie. the French and Spanish).

    Could their have been two waves of R1B? The first bringing the Basque and their kin into Western Europe and a second less than a thousand years later bringing r1B Indo-European Speakers who had picked up a West Asian component not present in their distant cousins, the Basque?

    Or the Indo-European invaders weren't R1B. But if not, what were they? Or maybe some other scenario? Like I said, I can't wrap my mind around it.
    Anybody else have any ideas?

    ReplyDelete
  8. Tracing the origins of the West-Asian Y-DNA component in modern Europe thus becomes key in further explaining the history of the Western World.

    My guess is we'll find that a large portion of that West-Asian component is sourced from the ancient Sumerians. I'm hoping we can find an ancient Sumerian corpse to test this on.

    I will disclose that my view comes from my idiosyncratic view of human origins. But this West-Asian late entry into Europe could well support my theory, if it is indeed Sumerian:

    http://richaelmussell.blogspot.com.au/2012/07/history-of-world-timeline.html

    ReplyDelete
  9. Thank you for this summary, it looks good and unbiased.

    There is still a question about mitochondrial haplogroups concerning the appearance and distribution of the hg H in Europe. It is quite widespread and would need explanation to have a big theory working. Without taking into account it we still lack one essential part of the big picture.

    ReplyDelete
  10. This PDF explains a little bit about the algorythm:

    http://dalexander.bol.ucla.edu/preprints/admixture-preprint.pdf

    I havnt read it, so I cant tell you if only math students understand it or if its really "basic". ;-)

    The algorythm bases on one, that students of mathematical statistic produced in the year 2000 and was first used in a software named "Structure".

    Admixture uses some kind of speed optimized version of this algorythm, wich resulted in identical results in 10% of the calculation time, wich seems still in the range of "many hours".

    --

    From my pure imagination, I expect Admixture to produce a set of SNP (genetical mutations) to define one "component".

    These dont seem to be exclusive, because components seem to have overlap, otherwise it wouldnt be possible to calculate "distances" between components.

    Also, sometimes Admixture comes up with components that are so exremely close to each others, that it virtually comes up with one drop rules like: your 100% this and the guy living next house is 100% that.

    Something like this show up if one tries to break Northwest european into more clusters.

    There you get one Irish centered cluster and one Swedish centered cluster. And, using national averages, they apear to losely resemble Celtic and Germanic. But Admixture seems unable to estaminate admixture between these 2 clusters but often comes up with a one drop. 100% Celt, 0% Germanic or 100% Germanic, 0% Celt. Wich is impossible.

    This makes me guess, the SNP lists for these 2 components have a huge overlap.

    BUt well, I dont know anyhting about it. You should better read the pdf. ^^

    ReplyDelete
  11. 'It reaches its lowest occurrence in areas where non-Indo-European languages have been spoken (Basques and Iberia in general, Sardinia, and Finland)'

    But the West Asian component also reaches its highest occurrence among non-Indoeuropean speakes: Abkhazians and Georgians..

    ReplyDelete
  12. But the West Asian component also reaches its highest occurrence among non-Indoeuropean speakes: Abkhazians and Georgians..

    Yes, of course, that's why I wrote:

    "the Indo-European speaking nucleus, originally one among many linguistic groups of the prehistoric Near East "

    A good way of looking at it is the following:

    Consider white North Americans: they have a lot of North_European. This North_European came mostly from British, Irish, Dutch and the like, and mostly with the English language, even though the groups with the highest North_European in Europe are Lithuanians, Russians, etc.

    Similarly, Spaniards brought Spanish and "Atlantic_Med" to South America, even though Sardinians and Basques have more of the "Atlantic_Med" there.


    Of course we don't know how the situation was in West Asia 5,000 years ago, but we cannot expect that the Proto-Indo-Europeans were necessarily the "champions of the West_Asian charts".

    ReplyDelete
  13. Thanks for your reply, yes it makes sense.

    Would this still live open the question of original admixture of the first Indoeuropean speakers?

    ReplyDelete
  14. Would this still live open the question of original admixture of the first Indoeuropean speakers?

    All we can say at the moment is that their gene pool probably intersected to a substantial degree with the West_Asian component. There may have been other elements in it that were not captured by the West_Asian component, but we can exlude most of them by a process of elimmination, e.g., they cannot have come out from territories where there was substantial "Southern" influence, because "Southern" is lacking in Indo-Aryans. Conversely, they cannot have come out from territories with "South_Asian" influence, because "South_Asian" is lacking in Europeans.

    ReplyDelete
  15. "the Indo-European speaking nucleus, originally one among many linguistic groups of the prehistoric Near East..."

    The problem with that is that you continue to rely on fringe linguistic theories and pseudo-linguistic statistics, while retaining the label Indo-European. Very few if any Indo-Europeanists support Dolgopolsky's belief that IE and Afroasiatic are related. You will be better off dropping the "Indo-European" label altogether and speak about the genetic populations illuminated by your analysis.

    Indo-Europeans, as a linguistic communioty, likely expanded from north of the Black Sea at roughly 3600 BC. Genetically, they may have had a West Asian component through pre-existing admixture from agricultural expansions originating in the Near East, but they came from Eastern Europe, not West Asia.

    "U-dominated older mtDNA stratum in Central/North-eastern Europe can be reasonably extended to cover both North-western Europe and northern Eurasia up to Lake Baikal, the prehistoric limit between Caucasoids and Mongoloids."

    Yes, this seems to be the situation, and this has been "known," without the U label, for decades. The Uralic language family is the closest in providing the geographic match for this trans-Eurasian linguistic unity. And as Saami demonstrate, there may be a special association between hg U and proto-Uralic.

    On the east, this Paleolithic population was replaced/admixed with the carriers of Y-DNA hg C and mtDNA A, C, D, G, Y, Z, all of northeast Asian provenance.

    Hg U seems to correspond to Y-DNA hg N as another pan-Northern Eurasian marker displaced by R in the west and C in the east.

    ReplyDelete
  16. The problem with that is that you continue to rely on fringe linguistic theories and pseudo-linguistic statistics, while retaining the label Indo-European. Very few if any Indo-Europeanists support Dolgopolsky's belief that IE and Afroasiatic are related. You will be better off dropping the "Indo-European" label altogether and speak about the genetic populations illuminated by your analysis.

    What utter nonsense! Did you even bother to read the post you are responding to? The argument is purely genetic.

    Indo-Europeans, as a linguistic communioty, likely expanded from north of the Black Sea at roughly 3600 BC. Genetically, they may have had a West Asian component through pre-existing admixture from agricultural expansions originating in the Near East, but they came from Eastern Europe, not West Asia.

    Those who believe in this theory (e.g., James Mallory) go at great lengths to deny influence from Near Eastern agricultural populations in the steppe. It seems we are making progress.

    Yes, this seems to be the situation, and this has been "known," without the U label, for decades. The Uralic language family is the closest in providing the geographic match for this trans-Eurasian linguistic unity. And as Saami demonstrate, there may be a special association between hg U and proto-Uralic.

    Could very well be, but that's not helping your case.

    Hg U seems to correspond to Y-DNA hg N as another pan-Northern Eurasian marker displaced by R in the west and C in the east.

    Absolutely no link. N is absent in most of Europe where the U stratum exists. N is unrelated to the U substratum; it came from the East and was added onto that stratum in the northernmost area of the boreal zone.

    ReplyDelete
  17. "What utter nonsense! Did you even bother to read the post you are responding to? The argument is purely genetic."

    Then don't use the linguistic label - "Indo-European." Indo-European is a category defined on linguistic grounds. You're just hijacking it to give a recognizable public meaning to your genetic "components" as in "The Bronze Age Indo-European invasion of Europe." If you want to understand the origins of Indo-Europeans, you need to build a solid linguistic argument, and then add a genetic argument to it.

    "Could very well be, but that's not helping your case."

    What case? You are the one building cases. Out of thin air, sometimes.

    "Absolutely no link. N is absent in most of Europe where the U stratum exists."

    N-M178 is found at 30-60% of Saami, Finns, Latvians, Lithuanians, Estonians. Exactly in the region where you have a mtDNA U-bubble in contemporary populations. It may have existed in wider Europe but was replaced by R-people. It's corresponding mtDNA U substratum survived better.

    ReplyDelete
  18. Mixing of the ANI and ASI did not happen 140 generations before as was believed, but probably more than 500 generations back (12,000yrs).

    Metspalu et al. (2011)

    ReplyDelete
  19. Excellent work D. Thanks for organising all this information into a coherent and clear picture.

    It might be worth looking into the IE religions for indications of IE ideology and social organization. The elite ethnic IE caste system in India is well known because of its continued existence and because of the Vedas. There are some studies out there of more general IE ideology and religion, Greek, Roman, Slavic etc. I will see what I can dig up.

    ReplyDelete
  20. N came to Europe from Asia, this is well known, and is also much more parsimonious than your alternative scenario whereby it represented an older stratum that survives only in NE Europe, was replaced everywhere else, but is surprisingly absent in ancient DNA samples from Europe.

    Of course, I could argue parsimony with you, but it would be a wasted effort on someone who thinks modern humans originated in the Americas.

    ReplyDelete
  21. Repo Man:

    "Since it is looking as if R1B arrived late, during the bronze age with the Bell-Beaker culture, R1B is the most likely candidate signal for the Indo-European invaders."

    I think Bell Beaker is a very unlikely candidate. They seem to spread from Iberia to NW and Central Europe, which is not impossible, but not likely way of the IE "invaders", but more importantly, there are no know fact supporting this idea.

    Also I think they are too early for that and IE reached West Europe only later.

    My personal opinion is that Bell Beaker was local, developed from the Iberian Megalith and R1b is older in West Europe than Bell Beaker.

    Which Hg Y is from IE movement, this is a long debated question here. Maybe they were very inhomogeneous for Hg Y that makes this difficult.

    ReplyDelete
  22. Since nobody makes new mtDNA maps these days, I fixed us one, from U5 data I found on a table:

    http://www.freeimagehosting.net/zy6v3

    ReplyDelete
  23. "N came to Europe from Asia, this is well known, and is also much more parsimonious than your alternative scenario whereby it represented an older stratum that survives only in NE Europe, was replaced everywhere else, but is surprisingly absent in ancient DNA samples from Europe."

    Hg U must have come from Asia, too. That's where all the R lineages are. Where else could it come from? From Africa? No. From Neandertals? No.

    Regarding ancient Y-DNA samples from Europe, we shall see. But note the following structural pattern: mtDNA U is "rare" in Eastern Siberia, while Y-DNA N is "rare" in Western Europe. This is a good example of differential, sex-biased lineage survival suggesting that Western Europe was colonized via a more male-dominant migration. mtDNA U and Y-DNA N overlap in between, which is, again, consistent with the retention by Uralic-speaking peoples of some of the essential genetic properties of that trans-Eurasian substratum.

    "I could argue parsimony with you, but it would be a wasted effort on someone who thinks modern humans originated in the Americas."

    Speaking about it, where's the American Indian component detected by multiple autosomal runs as also ranging from Eastern Siberia to West Asia?
    See the latest admixture graph http://blogs.discovermagazine.com/gnxp/2012/06/population-replacement-in-neolithic-spain/

    ReplyDelete
  24. Hg U must have come from Asia, too. That's where all the R lineages are. Where else could it come from? From Africa? No. From Neandertals? No.

    The difference being that mtDNA haplogroup U's earliest ancestor is one order of magnitude older than the earliest ancestor of European N-Tat.

    But note the following structural pattern: mtDNA U is "rare" in Eastern Siberia, while Y-DNA N is "rare" in Western Europe.

    The pattern is not that difficult to explain, given that mt-haplogroup U originated in Western Eurasia and Y-haplogroup N in Eastern Eurasia.

    Speaking about it, where's the American Indian component detected by multiple autosomal runs as also ranging from Eastern Siberia to West Asia?

    Remember that your Homo Sapiens Out of America theory is verboten here.

    ReplyDelete
  25. Repo Man
    "Since it is looking as if R1B arrived late, during the bronze age with the Bell-Beaker culture, R1B is the most likely candidate signal for the Indo-European invaders."

    I think it's the other way round. I think the original clade of R1b was introduced by farmers coming round the atlantic coast and the spread of the specific clades associated with western europe was the result of the spread of a foraging / fishing / cattle raising culture associated with megalithism which happened to suit the climate. So a pre-IE (or southern wave IE) west to east expansion with the IE (or northern wave IE) coming later.

    ###

    If the first farmers were in eastern Anatolia and the Minoans, Etruscans, Sumerians etc were non-IE and they were displaced by IE speakers then it does seem most likely to me that IE *originated* in the mountainous regions to the north and east of Anatolia with offshoots to the balkans and the steppe.

    It also seems likely to me that both the origin and the offshoots could have been the source for different expansions at different times with the steppe one - the one people usually associate with IE - coming last.

    So the sequence (maybe?)
    1) non-IE first farmers in eastern anatolia spreading west and south
    2) PIE in mountains north and east of anatolia adopt pastoralism under the influence of the first farmers
    3) PIE split
    - one branch to Anatolia
    - one branch to the balkans (becoming the new main source)
    - one branch to the steppe (becoming a new main source)
    4) balkan IE split
    - one branch follows the non-IE speakers west via Greece and the med (relatively rapidly)
    - one branch expands up the danube (relatively slowly)

    [optional
    5) west to east / south expansion of atlantic cowboys after the introduction of cattle]

    6) north-eastern IE expansion (now majority north european through mixture) from the steppe pushing south and west which is the one people generally associate with IE.

    So two IE expansions into Europe with different genetic signatures: an early, rapid mostly west asian southern one heading east to west followed by a much slower northwards expansion limited by latitude and a later, mostly northern euro, north-eastern one heading west and south.

    (Separately
    3b) Semitic tribes to the south of Sumer do the same thing i.e. adopt pastoralism from the Sumerian farmers and eventually over-run them.)

    ReplyDelete
  26. Is it likely that expansion of groups from the Black Sea area might have brought the Indo-Europeans languages to Europe to such an extent that probably most of the former and potentially unrelated languages went extinct with few remaining traces (excluding some languages)? However, I find a combination of the two major theories (farmers vs animal husbandry) about the origin of the Indo-European language family more convincing, and compatible with the genetic data, than either one of them alone. It is conceivable that the widespread occurrence of Indo-European languages already in pre-roman times is chiefly due to migration of farmers associated with demic diffusion. Secondly, farmers on the boundary of the steppes may have developed a pastoral lifestyle. Somewhat later, perhaps in the Bronze Age, these semi-nomadic communities might have pushed into central Europe, either forcing other Indo-European speaking groups to migrate or implementing language shift where they settled down. Thus, through these mechanisms, the result would be an expanding range of people speaking Indo-European languages.

    ReplyDelete
  27. Could you run ADMIXTURE with only the ancient samples excluding modern samples?

    ReplyDelete
  28. The ancient samples have a limited number of SNPs and once you intersect all individuals you're left with pretty much nothing; that's why it makes little sense to run ADMIXTURE with them alone.

    ReplyDelete
  29. @Atle OStern
    Hi, your theory had already been proposed by Hard Goodenough, in his book "The Evolution of Pastoralism and Indo-European Origins"

    Thus, the majority of scholars see the introduction of Indo-European languages to western Europe as coming after the spread of agriculture in a piecemeal process starting before 3000 BCE and continuing until the present. An eclectic hypothesis I see no reason why the hypotheses of Gamkrelidze and Ivanov, Renfrew and Gimbutas cannot be reconciled or fruitfully combined. We all accept the origin of Indo-Hittite in Anatolia, against the traditional vision of an Urheimat in the steppe. Where Gamkrelidze and Ivanov see “Greek” as having moved across the Aegean, however, I agree with Renfrew that the initial move was made much earlier with the spread of agriculture. I differ from Renfrew in seeing the migrants’ language not as Proto-Greek but as a branch of Indo-Hittite. Peoples speaking forms of this language spread north to create the Neolithic civilizations of the Sixth and Fifth Millennia BCE in the Balkans. Here I differ from Gamkrelidze, Ivanov and Gimbutas. I then follow the scheme set out by W. H. Goodenough in 1970. He argued that people from these agricultural civilizations on the edge of the steppe developed techniques of nomadism. From this mixed agricultural and nomadic population that spoke Indo-Hittite the Kurgan culture formed and Indo-European, in the narrow sense, developed in the Fourth Millennium.27 At this point, I accept the conventional view that the Kurgan culture and Indo-European languages spread out from the steppe. What Gamkrelidze and Ivanov call the Ancient European Dialects (Celtic, Italic, Illyrian, Germanic, Baltic Slavic and, probably, the Tocharian families) derived from northern dialects and migrated earlier, while the Indo-Aryan (Armenian and Greek) came from the southern. It seems that Indo-Iranian speakers had penetrated Iran from the north by the end of the Third Millennium BCE. During the Second Millennium, they entered the Near East and conquered much of northern India. Already they appear to have been calling themselves Arya or Aryans.

    ReplyDelete
  30. From my alternatively-biased point of view:

    We have a mesolithic Hunter-Gatherer in the La Brava samples.

    Two probable hunter-gathers from the transitional period in the AJ samples.

    One definite hunter-gather from transitional period in Oetzi (I could accept that he might be from farming stock, but that is not the life he led).

    I am not sure about GOK but farmer for the sake of argument.

    The oldest samples, La Brava Mesolithic Hunter Gatherers do indeed look like a mating between the AJ like folk and the GOK/Oetzi folk. This suggests to me that both groups may have been around in the mesolithic. Which makes them both pre neolithic.

    I too am curious about the EAst African. I could believe West African (via Gibralter) more easily. East Africa seems a long way away.

    ReplyDelete
  31. I think it wasn't a good idea to make the composite individual from La Brana samples, for they are not very homogenous.

    Although my results are somehow similar to yours, i've discerned a visually observed pattern in the smooth gradient of North European mesolithic genetic component.

    According to the Admixture K=4 results, the Mesolithic component is modal in BRA1, immediately followed by Latvians,Lithuanians,Estonians, Karelians, Belorussians, Finns and Samis.

    On other side of the spectrum we have Cypriots, Öetzi, Gök4 and Ste7 - Neolithic farmers who have 0% of North European component.

    http://image-upload.de/image/bUVw9q/179189e059.png

    ReplyDelete
  32. Correction:
    BRA1 is followed by AJV52 and only then by Latvians

    ReplyDelete
  33. "One definite hunter-gather from transitional period in Oetzi (I could accept that he might be from farming stock, but that is not the life he led)."

    If he wasnt Farmer, then he definately traded with Farmers, because he clearly had Farmer produced products with his corpse.

    Maybe he is something like an American "trapper".

    "Trappers" came from Farmer (european) stock, not from HG Stock (native North-Americans), but lived the life of a Hunter Gatherer, with Farmer produced weapons and tools.

    ReplyDelete
  34. I see no contradiction originating from a farming community and trading with farming products or products that are produced within a farming community. For instance my Grandfather was a landowner but didn’t till the soil himself, though he was a merchant for agricultural commodities. I wouldn’t be surprised if Ötzis likely higher status within his farming community got him killed.

    ReplyDelete
  35. Hello,


    First of all I would like to thank the moderator and author of the study, for the effort, and really for opening up such a debate.

    Secondly, I speak as no expert in the domain, but as a passionate of the historic truth and knowledge about our ancestors. With an expertise in Sociology & Communication studies, during a private academic research period in Eastern Europe, I got the chance to explore the regions rich folkloric heritage and, through Romanian friends, I also received alot of information regarding an important debate about ancient history of Romania and of their nation. Entire episodes in local pre-history and history are revealed as missinterpreted or as being on purpose hidden or missconstruded ( archeological evidence and scientific reasoning contradict established history ).

    Regarding the IE debate, and their DNA traces in modern European populations; has anyone really followed the linguistic connection between ancient european languages, PROTO-Thracian/Dacian, Pelasgian traces, in conconrdance with the IE archaic languages ? This connection might also find its value in genetic studies, as populations close to the Black Sea shore might be able to reveal more information about migration routes, time period, and overall DNA mutation traces. Furthermore, remains found in regions of what is now Romania and Moldavia seem to point out to local archeological and antrhopological disputes ( Romanian history) regarding ancient writings in clay that resemble ancient summerian/asyrian/proto-greak writings. Also, ancient Paleolithic human remains have been uncovered in cave systems such as the the cave known in local Romanian folklore as "Pestera cu Oase" / "en. The Bone Cave". Even though local Romanian research is ideologically biased and financially hindered, such facts shoudl not ignored.

    It is my firm belief that further study on ancient Romanian and overall Pontic populations might be able to reveal interesting connections and information regarding our early human history. Such scientific endeavours are still in infancy in the Eastern European region, but still interest exists and there is an abundance of local help and resources that can be accessed..

    Also, such a pluri-disciplinary approach would be able to exemplify for a wider picture of the domain, at least helping to back-up data with scientific reasoning..

    I woud very much want to thank anyone who'd spare the time to give me any kind of feedback regarding such approach. Also, many thanks for the opportunity to get in contact and share information.

    Respectfully,
    Razvan Bera

    ReplyDelete

Stay on topic. Be polite. Use facts and arguments. Be Brief. Do not post back to back comments in the same thread, unless you absolutely have to. Don't quote excessively. Google before you ask.